As a property professional, chances are you have already encountered a complaint that looks rather different from those you were receiving a few years ago. What was once a relatively straightforward email may now arrive as a lengthy, highly structured argument, complete with technical terminology, legal references and a whole host of issues that may or may not be relevant. Sound familiar?
AI is changing the way consumers complain and, for agents, that can mean more time spent unpicking arguments before getting to the heart of what has actually happened. I’m afraid it is a topic we cannot get away from, so the best thing to do is address it, embrace it and learn from it.
At Property Redress, we are certainly seeing AI become very common. Anecdotally, I would say around 85% of the consumer complaints we now see have some element of AI involved, usually where somebody has used it to help formulate or structure their argument.
I am not suggesting this is necessarily a bad thing, AI is a reality of the world we live in and, used properly, it can be incredibly useful. Consumers will often turn to AI because they are trying to understand whether what has happened to them is normal or reasonable, what industry best practice looks like and whether something has gone wrong. That can be genuinely helpful, particularly for somebody who knows very little about the property industry and is struggling to articulate the problem they have experienced.
The difficulty, however, is that AI can identify every conceivable issue and suddenly what might have been a relatively straightforward complaint becomes much bigger and more complicated. It does not necessarily understand which points really matter, what is reasonable in the circumstances or how expectations should be managed.
At its best, AI can simplify a complex situation. At its worst, it can send somebody down a rabbit hole of issues that are not necessarily relevant to resolving their complaint.
When a complaint becomes a 10-page argument
We are increasingly seeing complaints that are far more technical and complicated than they need to be. They may be very well structured, but that does not always mean everything in them is correct or necessary. These complaints can also lose some of the emotion and human element you would normally expect and, increasingly, they all start to sound quite similar.
I recently dealt with an example where somebody complained that their agent had not provided the ‘grantor’s information’. Initially, it was not even clear what information they were asking for or why. Once we queried this, it turned out they were the freeholder and simply wanted certain information relating to payments. Somewhere along the line, they had come across a technical term, put the issue into AI, and it had built an argument around it.
That is where AI can sometimes be unhelpful. Introducing technical terminology or legal references does not necessarily make a complaint stronger. In some cases, it can actually make it harder to establish what has happened and what the consumer wants the agent to do about it.
For agents receiving these sorts of complaints, my advice would be not to get too distracted by the length of the submission or the legal terminology being used. Go back to the basics. What is the consumer actually complaining about? What happened? What evidence do you or they have, and what can reasonably be done to resolve it?
The team and I increasingly find ourselves wading through 10 pages of legal references and arguments before getting to those fundamental questions. I often use the analogy of standing in front of a judge. There is little value in spending 15 minutes explaining the law to a judge who has spent years studying and applying it. What they need from you are the facts of your particular case. It is similar with redress.
Winning an argument is not the same as resolving a complaint
Our approach to complaints and redress is deliberately non-adversarial. We are trying to establish what has happened and find a fair and practical resolution. AI does not necessarily understand that. If you ask it to help you make your case, it can approach the situation as though it is there to win an argument rather than help resolve a problem.
The danger is that somebody can become convinced they have a cast-iron case because of the way AI has presented it to them, when the practical circumstances or evidence simply do not stack up in the same way.
However, I think it is really important to distinguish between an AI-generated complaint and a vexatious complaint, because they are not the same thing.
At Property Redress, we have actually removed the term ‘vexatious’ from our Terms of Reference. Complaints often come with a lot of emotion and anger, but in most cases people are not deliberately trying to play the system, so we need to be careful about applying labels.
Whether an agent believes a complaint is vexatious or not, our job is to consider the facts and review it independently. A complaint cannot be upheld without evidence. AI can help somebody make an argument, but it cannot substitute for the evidence needed to substantiate it.
What should agents be doing?
For agents, the challenge is not simply how to respond to AI-generated complaints, but how to make sure their own complaints procedures are robust enough to deal with them. Some of the biggest traps remain the same as they have always been, namely failing to recognise a complaint early enough and insufficient record-keeping.
A complaint is not necessarily a reflection of what you do day to day. If you are doing most things well, nine times out of ten a consumer will not thank you for it. However, when something goes wrong, you can guarantee they will come back to you if you do not step up and deal with it.
Complaints give you useful information about where things can go wrong. The complaints process is not about winning an argument but resolving the issue and considering whether anything needs to change internally.
Similarly, agents should not be afraid of using AI themselves. It can be extremely useful for summarising information, organising evidence and timelines, and helping prepare responses, but without doubt, you still need humans at the heart of the process.
My biggest warning, however, is around data. If you are going to use AI, make sure you have an appropriately licensed system that your firm controls and understand exactly what happens to the information you put into it. You cannot simply feed personal or identifiable information into AI without considering where that data is going and how secure it is. Firms need proper training, protocols and guidance around its use. In my view, as AI becomes more common, organisations and regulators should also be considering whether their guidance needs updating.
Human first, AI second
AI-generated complaints are here to stay, and there is nothing we can do to stop that. Nor should we necessarily want to.
AI can be incredibly useful, but it does not have empathy, lived experience or a complete understanding of the circumstances surrounding a complaint. It can support the process, but it cannot replace human judgement.
At Property Redress, regardless of what is put forward to us, we will always try, and mostly succeed, in speaking to both parties and asking them to explain to a real person what has happened and what they actually want. There is a reason for that.
In the end, I think the overarching message is around the importance of authenticity and remembering that there is a person behind every complaint. They may have used AI to construct a lengthy, technical argument, but ultimately, they are somebody who feels let down by the service they have received and wants their concerns to be heard.
For agents, the challenge is not to get caught up in fighting every argument AI has helped produce, but to use your own human skills to understand what has actually gone wrong, defuse the situation and find a fair resolution.
We must not lose sight of the fact that we work in a service industry, and good service requires human interaction, empathy and understanding. For me, the key is human first, AI second. AI is a tool and essentially a servant. It is there to support the process, not replace the human judgement and personal interaction that are so important when resolving complaints.
Sean Hooker is head of Redress at Property Redress.
